Literature study
A study on: Online analysis of photo-degradation
By: Mohamad Ahmad
Student number: 11355662
Supervised by: Freek Ariese
2
ndReviewer: Govert Somsen
October 25
th, 2018
Table of Contents
Summary ... 3 1 Introduction ... 4 2 Theory ... 6 2.1 Samples ... 6 2.2 Photo-degradation ... 6 2.3 Separation techniques ... 9 2.4 Photo-reactor ... 10 2.5 Detection techniques ... 11 2.6 Hyphenation ... 123 Review on on-line analysis & photolysis ... 13
3.1 Without separation – on-/off-line ... 13
3.2 Off-line separation ... 33
3.3 On-line separation ... 54
4 Conclusion & Discussion ... 69
5 References ... 70
Summary
A review has been performed on the procedures applied for the on-line analysis on photochemical processes. A broad overview is given on the various aspects that come into play, looking into the strategies that involve Off- and On-line procedures, whereby the photochemical processes are either connected or separated from the analyses, giving one the opportunity to either monitor the processes “live” (On-line) or investigate the long term effects (Off-line). Added to this is looking towards incorporating chromatographic methodologies to enhance the overall analysis. An investigation was done on the different kinds of photochemical processes that occur and the chemical and analytical tools that accompany it, while giving an overview on the various obstacles one might come across during such investigations. A final conclusion is given on the theoretical development of a truly comprehensive methodology, giving the researcher great control on the overall processes.
1 Introduction
Photo-degradation[1] (PD) (Chapter 2.2) is the alteration of a molecule by means of light absorption.
This field of study is massive, consisting of many subsections. In most cases, an investigation on photo-degradation consists of tackling only one or more branches of the overall subject. PD plays a major role in many fields, of which arguably the most known is the conservation field[2-6]. As it is crucial in the
preservation and restoration process of historical objects, these being, paintings, sculptures or other artworks. These objects are subjected to various environments that could degrade it, e.g. by oxidation, heat, light or a combination of those factors[7,8]. Of which, light is a crucial point of investigation. PD is
not only observed is artistic works, but is also seen in agriculture (pesticides) [9-12], pharmaceuticals
(drugs)[7,13-20], plastics (polymers)[8, 21-24], foods[11,25] and (waste-) water[14,19,26-28]. Although PD is
portrayed in a negative light, with respect to artistic works, as it alters the original work, within other fields PD is used as a tool, to assist or improve on the procedures applied. For example, in agriculture, there has been a rise of concern on pesticides in foods, as such there is an investigation on the PD pathways of said pesticides, to convert to “innocent” non-toxic molecules, preferably minerals[10]. In
the wastewater treatment field it is utilised as a destruction procedure, remove any toxic molecules in the water, changing them to non-toxic compounds, creating safe drinking water.
Although these points are proven to be true, whereby PD is used as a purification tool, this point has another side to it, which is the contamination side of PD. This is seen in pharmaceuticals, plastics, agriculture and water, whereby the PD leads to an increase in toxicity, that was initially either non-existent or insignificant. Within all of these field, it can lead to the production of highly toxic molecules that can affect the environment (human, animal or plant life). For example, river water, which in most cases already contain an abundance of molecules, can under the effect of constant sunlight produce highly toxic chemicals that effect the surrounding environment. Added to this, the constant insertion of pollutants by humans, which can consist of plastics, pharmaceuticals or other kinds of waste, can have an immense impact on the environment. A more direct effect of PD on humans is with foods, plastics and pharmaceuticals. Beer, for example, changes the flavour if under the exposure of UV-light, as such, most manufacturers utilise glass, to block out this light. With pharmaceuticals, this becomes dangerous, as drugs already have to purpose to affect human life, with the PD of these drugs, it can either deactivate the drugs or altering the molecular composition is such a way that the effect becomes unknown with the possibility of becoming toxic. Similar to drugs, plastics can also have a negative impact, as the PD of these materials can produce toxic chemicals, with for example drinking bottles or plastic wrappings. The PD products can then be accidentally ingested, and effect a person in a negative way.
The examples mentioned show that PD is a very wide subject, covering various fields, seen in many different kinds of matrices. As such, the investigation of this process is also very wide and is in most cases very complex, taking river water as an example, these kind of samples can contain many different bacteria, particles, proteins, drugs and other types of molecules. Due to this complexity, the need for a more comprehensive technique increases. Observing particular PD processes becomes nearly impossible for complex samples, if the compound of interest is not properly isolated, as it can be affected by the various molecules that it comes in contact with (see chapter 2.1). This interaction with other molecules can inhibit or enhance the PD process and give skewed results. For a proper determination of specific PD process, a more comprehensive technique is necessary. Whereby a full investigation can be made on a specific sample. This technique would consist of a separation method that would isolate specific compounds, after which the isolated compounds undergo a PD procedure. The products from this procedure would then be separated and detected, to acquire the information on the precursors (initial compounds) and products. Such a technique is as of yet not developed.
The goal of this literature study is, as such, the investigation of the PD procedure and analysis. Looking towards the current procedures applied in the various fields. Emphasising on the analyses, by going over the various points of the apparatuses and procedures applied in various studies. In chapter 2, the theoretical background necessary will be explained, going over the PD process, the various conditions that come along, the samples / molecules that are investigated and the various machinery that is utilised. Chapter 3 will contain review on the various literature that is investigated, however this chapter will split into three subsections, discussing various types of analysis. Starting off with Chapter 3.1, which goes over the analysis of the PD process by utilising only detection techniques, whereby no separation of the various compounds in the sample is applied. The analyses discussed however, will go over on-/off-line procedures applied. These analyses are either with the PD process attached to the detection or separated from the analyses altogether. Chapter 3.2 will go over the analysis, whereby a separation technique is utilised after the PD procedure is applied (off-line), to determine the PD products, as well as giving an indication on the kinetics (see chapter 2.2) that come with in. In the last subsection, Chapter 3.3, the studies on the PD process utilising the separation techniques with a PD procedure in an on-line manner will be discussed. This technique will go over the utilisation of a photo-reactor (see chapter 2.4) to connect a separation apparatus with a PD procedure. In chapter 4 a conclusion will be given on the various studies discussed with some points on the application of a fully comprehensive technique that has a mentioned before, never been seen in the current literature. The scope of the thesis is to investigate the methodology of analyses on the PD processes, giving only a surface level understanding on PD itself.
2 Theory
This chapter will discuss the theoretical background.
2.1 Samples
As mentioned in Chapter 1, PD is seen across many fields of study and as such is seen in a variety of environments, whereby many different kinds of molecules are investigated. Some generalisation is applicable, as in most of the studies investigated, the molecules are soluble in water. Narrowing it down even more, to organic molecules that contain carbon, hydrogen, nitrogen and oxygen atoms. This is even more so the case with studies that looked into the investigation of direct photolysis (see Chapter 2.2), as the molecules have to absorb the emitter light, of which it will range from deep-UV to near-infrared light. Only organic molecules will modify will these energy levels, as other types, such as minerals and salts need higher energy levels to achieve any alteration. The altered molecules (after PD) will split and will then be analysed with a detection technique. In scheme 1 a simple PD process of a N-Nitrosodialkylamine is shown, whereby with the irradiation with 366 nm light a NO group splits off of the original compound. After some rearrangement, the products become a RN (H+) RHC and HNO compound. This is only one specific example, in the studies investigated different conditions with different molecules will be shown.
Scheme 1:PD process of N-Nitrosodialkylamine
The molecules investigated come from pharmaceuticals, wastewater, paints, agriculture and some other miscellaneous fields. With these fields of study specific matrices in which the molecules are located can be speculated. For example, drugs from the pharmaceutical field will most likely be in matrices that are, blood, for testing (on life), pure water for analysis and wastewater for environmental testing. For wastewater, it will be as the name suggests, in matrices like river water or water from water treatment plants. For paintings it will be oils and minerals, as paints most likely consists of these types of molecules. Lastly, for agriculture, it will be on the subject of pesticides, as such could be within matrices like foods, beverages or soil. However for all of these (except for pure water) matrices some form of filtration or purification will be necessary. Most of the matrices discussed could contain particulate (solid) matter, high concentration of proteins, compounds similar to the molecule of interest or other compounds that could affect the PD process in an unknown way. With the use of a separation methodology before PD, the molecule of interest can be isolated, this however is further discussed in Chapter 3.3.
2.2 Photo-degradation
Photo-degradation is the modification of a molecule either directly or indirectly affected by light. Degradation, in this instance, means a (significant) change of the molecule and may not necessarily mean destruction of the molecule, but could also mean, for example, addition, dimerization, isomerization, alkylation, etcetera[1]. The light necessary for modification is usually in the ultraviolet
and visible (UV/VIS) range, because the changes that lower energy (≤ infrared) light applies to a molecule is usually insufficient for any modifications to happen, although there are studies done on degradation with infrared light, this is usually done with the aid of many external elements[43], e.g. using
a catalyst, or by applying high intensity IR light to only a single molecule. UV/VIS light does provide sufficient energy to modify the molecule for degradation, this being excitation of electrons. Another justification for the use is the fact that (natural) sunlight consists for the most part of the UV/VIS range[29], sunlight does not contain wavelengths of light below this (< deep UV) as it is absorbed by the
atmosphere. The range of sunlight (at sea level) in terms of wavelength is around 300 to 1600 nm, excluding the low intensity (≤ infrared) light, see figure 1 for the solar spectrum.
Figure 1: Solar spectrum at different levels (taken from [29])
If anything above the longest wavelength of visible light is removed, a range of 280 to 760 nm is left. ISO 21348[30] defines the solar spectrum in (sub-) classes (see appendix 1). The classes UV (UVA, UVB,
and UVC) and VIS are used in most research projects[10,13,16,31]. These abbreviations are usually used if
sunlight is either used or artificially recreated. These ranges of wavelengths can cause certain electronic transitions to happen in a particular molecule, this differs per molecule. This absorption of light is quantised, as such only specific wavelengths of light will be absorbed by specific molecules. The transitions π π* and n π* are generally reported in the range of 200 to 750 nm, while σσ* transitions only occur below 200 nm[32]. These transitions will be the main cause of either direct or
indirect photo-degradation. Direct meaning the particular molecule is excited by light directly undergoing the specified transitions after which, the modification occurs. With indirect the meaning is that light is absorbed by a sensitizer after which it interacts with the compound of interest. Although direct absorption of light quite simple and straightforward, it can still be affected by nearby molecules. This lies with the “competition” of absorption. Due to the finite amount of photons that is beamed at a specific sample, if multiple compounds absorb at the same wavelength, not every molecule will become excited. This will depend on the molar absorptivity, which is different for each molecule. For indirect PD, a nearby molecule absorbs the light after which it can follow different pathways. These being, the molecule modifies to create in most cases a radical, which in turn reacts with the molecule of interest. A radical is a molecule with an unpaired valance electron. The final path is that a nearby molecule absorbs the light and either emits (fluorescence) a photon of light or affects the environment to form
reactive species that interact with the molecule of interest. The most common reactive species are produced from water and oxygen, as they are in most cases the most abundant molecules seen. In a reaction with dissolved oxygen in water and light (< 310 nm), singlet oxygen and hydroxyl radicals are formed, which in turn react with the molecule of interest[33,45].
From this the importance of knowing the environment of a sample becomes apparent. For example, in paintings a known mineral utilised in white paint is TiO2[3]. The mineral was used due to its very white
colouring, this mineral however has a drawback, which is the fact that the mineral increases the speed of degradation. This is due to the fact that it catalyses the production of hydroxyl radicals, which in turn will react with its environment, degrading it. After investigation, an interesting property was found, it would seem that not all forms of the mineral would show this catalytic behaviour. The two most known forms are Anatase and Rutile, of which Anatase is the catalyst for hydroxyl formation. The Rutile form would absorb light, but not be a catalyst, while the Anatase form would increase the production of hydroxyl radicals. The Anatase form could be seen as enhancing PD while Rutile could be seen as either inhibiting (due to its absorbance of light) or being non-existent to the PD process. The reasoning behind this difference is outside of the scope of the thesis and as such, will not be discussed. In figure 2, a schematic illustration of the process is shown.
Figure 2:A schematic illustration of the ongoing processes near TiO2 particles (taken from [34])
Another important point that is connected with this, is the kinetics that come to play in the PD process. Kinetics is another word for reaction rate, which translates to the speed of a specific chemical reaction. This is in most cases portrayed as k with the unit being s-1, giving an indication on the speed. The general
formula for first order reaction is seen as:
[𝑨]𝒕= [𝑨]𝟎∗ 𝒆−𝒌𝒕 (1)
Where [A]t is the concentration of the molecule of interest at time t (or 0) and t equates to time (seconds).
A negative sign (-) is next to k, due to the fact that PD is a “destruction” process, where A is removed, and as such the concentration will decrease. In figure 3 a simple kinetics plot is shown, whereby the concentration (mol L-1) is on the y-axis and time (min) is on the x-axis. Both the initial molecule (4-CA,
blue) as well as one of the products (CL-, red) are shown. If the natural logarithm is taken, a linear relation will form, of which the slope will become –k (for the PD of 4-CA). This k will represent all the processes that occur to degrade the 4-CA compound. From the graph one can see that the concentration of CL- at 120 minutes is not equal to the concentration of 4-CA at 0 minutes, meaning that there are one or more minor processes occurring. This specific example is further explained in Chapter 3.
Figure 3: Concentrations (mol L-1) of 4-CA and chloride as a function of photo-irradiation time
In this specific example the sample is irradiated with artificial solar radiation, without the assistance of other compounds or catalysts. As such, the kinetics plot can be “modified” by using these external tools, and although these kinetics plots shows one the processes going on, it does not give any in-depth information on the photo-physics. The term quantum yield on the other does give some insight into this matter. Quantum yield[14,33,44]is a general term used in photo-physics to give an indication on the number
of processes occurring per photon absorbed. This is a general term, as different quantum yields exist, two of those being, the quantum yield of fluorescence and decomposition. It translates to a fraction that has the unit of the number of events occurring per absorbed photon, shown as ɸ:
ɸ = # 𝒐𝒇 𝒆𝒗𝒆𝒏𝒕𝒔
# 𝒐𝒇 𝒂𝒃𝒔𝒐𝒓𝒃𝒆𝒅 𝒑𝒉𝒐𝒕𝒐𝒏𝒔 (2)
For each quantum yield, the events are changed, for fluorescence, this becomes the # of photons emitted, while for decomposition this becomes the # of molecules decomposed. White et al[39] showed an altered
formulae (3), whereby the two forms of quantum yield were combined to determine nf, which
corresponds to the total number of photons that a fluorescent molecule can absorb before decomposing.
𝒏𝒇=
ɸ𝒇𝒍𝒖𝒐𝒓𝒆𝒔𝒄𝒆𝒏𝒄𝒆
ɸ𝒅𝒆𝒄𝒐𝒎𝒑𝒐𝒔𝒊𝒕𝒊𝒐𝒏
(3)
The values of ɸ in most cases is between 0 and 1, which translates to either nothing happening with each absorbed photon, of something happening with each absorbed photon. A value of 0.5 would then mean that with each two photons absorbed something would occur, this could be either fluorescence, decomposition or another event entirely (not further discussed).
2.3 Separation techniques
Many kinds of separation techniques are used in the analysis of degradation products, with most research areas applying liquid/gas chromatography (LC/GC) to their analyses [6,7,9,19,35]. This is due to
the fact that most compounds that need to be analysed are organic molecules, containing carbon, hydrogen, oxygen and nitrogen atoms (see Chapter 2.1). The molecules, although containing much of the same atoms, do differ a lot in size. Ranging from small molecules [7] (100 Da) to polymers [21] (100
KDa). The term LC is very loosely used. When mentioning LC, this usually refers to reversed phase liquid chromatography (RPLC). Although this is used is most papers, this is not the only separation
technique applied. There has been research where, for example, ion chromatography[36] (IC) or size
exclusion chromatography[6] (SEC) was used for the separation of charged molecules or polymers,
respectively. LC is generally used, due to its versatility as it can separate a broad range of molecules, going from small molecules (hundreds Da) to polymers/proteins (100 KDa). Depending on the particular set-up that was used (column, pump, etc.). The scope of this study is limited to only LC, because of its use across most, if not all, areas discussed in this study. Therefore, only LC will be discussed further in this sub-chapter. LC is a technique used to separate mixtures of different compounds. A mixture is dissolved in a fluid, the mobile phase, which “carries” the mixture past a solid substance (usually beads) in a sealed environment (column). This solid substance is called the stationary phase. The different compounds in the mixture have different speeds, causing them to separate. The speed at which the compounds travel is based on differential partitioning between the mobile and stationary phases. If a certain compound partitions into the stationary phase, it remains in the same position until it partitions back into the mobile phase. The mobile phase moves at a certain constant speed, called the flow. This can be varied, to acquire different results. The partitioning coefficient depends on the properties of the substances and the mobile and stationary phase. There are 2 main LC techniques that are widely used, Normal (NPLC) and Reversed (RPLC) phase liquid chromatography. The latter is more used because of the mobile phases that are used. In RPLC the stationary phase is significantly less polar than the mobile phase and in NPLC this is the opposite. RPLC uses more polar mobile phases, which is ideal for organic compounds/matrices. This is also very advantageous for the various detection methods that will be discussed in Chapter 2.5.
2.4 Photo-reactor
The actual practical matter of applying light to the samples can quite differ between apparatuses, depending on the situation. Starting with the simplest set up, which is a UV-Chamber. This chamber consists of a light source, a place to position the sample and is usually closed off with reflecting materials (mirrors), as to keep the leaking of light to a minimum. This is an off-line approach to PD, as the sample is placed in the chamber and after irradiation removed again, to be analysed by whatever means are necessary. With all the photo-reactors discussed, they can be temperature and pressure regulated to keep the conditions stable, this regulation can be specific for the UV-chamber or can cover the room or area the UV-chamber is placed in. The next step in photo-reactors is utilising it in an on-line approach. Starting first with utilising it with only a detection system. The first and arguably most simple matter is utilising a spectroscopic technique, whereby the detection light source and the PD light source can pass through the sample at the same time. In such a way, that continuous detection can be utilised. This however, bring some issues, as the light source (or emission of sample) cannot overlap. The light sources should not affect the process of each other, meaning the PD light source should not interact with the detection approach. A more complicated approach is utilising MS at a detection technique. This is quite intuitive as for MS, ionised molecules are a necessity. The small organic radicals that are formed are much more easily ionised, as such one could expect an increase in signal, depending on the conditions. However the obstacle that needs to be overcome in utilising a photo-reactor in a continuous manner, meaning that a constant flow is needed. The same problem is seen when trying to combine a separation technique with a photo-reactor. The solution to this is the utilisation of PTFE-tubing. PTFE (polytetrafluoroethylene), also named Teflon, is a synthetic polymer that is very hydrophobic, does not degrade and has a low UV/VIS absorption. This material is perfect for PD approaches, as it does not interact with the environment at all. This tubing is utilised across most of the examined literature that utilises on-line PD. In one particular case a TiO2 coated glass tube was utilised
to also act as a catalyst[37]. However, for the other cases the PTFE-tubing was utilised, whereby the
tubing was coiled around the light source (see chapter 3.3 for illustrations). Due to the lack of absorbance of the tubing, the light source could just illuminate the tubing, for the PD process to initiate.
2.5 Detection techniques
There are various ways of detecting degradation products. The main approaches being either Mass Spectrometry[11,19,37] (MS) or Spectroscopy[4,23,35] (e.g. UV/Vis). Although NMR has been used [16], this
was only reported in some cases. In general, either MS or Spectroscopy was used. UV-Vis/Fluorescence techniques are generally reported to be used as detection methods when mentioning spectroscopic approaches. The justification for the use of these detection methods is that the compounds that are to be analysed are usually of an organic nature (see Chapter 2.1). These kinds of molecules are ideal for spectroscopic techniques and MS. Organic molecules are easily ionised in MS and usually have an absorbance spectrum for UV-Vis. There have been reports [28] of the use of both MS and a Diode Array
Detector (DAD). A DAD records the absorbance at multiple wavelengths (UV-Vis range) of light at the same time.
Some minor detection systems used will be briefly explained. Starting with X-ray photoelectron spectroscopy[3] (XPS), it is a technique whereby the surface of a sample is irradiated with X-rays (< 10
nm). This high energy irradiation causes the inner electrons of atoms to “shoot out” of the atom into a detector, which is usually positioned at an angle. The energy that is contained in the electrons is determined by the detector and due to the fact that the energy of the x-ray is known, the energy of the electron before the atom absorbs the x-ray can be determined. These energy levels are quantised within the atom as such, the atom can be determined with this information. A drawback of this technique is that it only can penetrate up to ~10 nm of the sample, as such the technique is mostly utilised in surface analysis. Another surface analysis technique is atomic force microscopy[3] (AFM), which can determine
the physical form of the surface. This is done by, to say so simply, put a pin on the surface and determine the height of the pin, moving along the surface and its grooves. This technique will only give information on the physical surface of the sample. The next technique is electron spin resonance[38]
(ESR) spectroscopy, this is very similar nuclear magnetic resonance (NMR), due to the fact that is studies the spin property. However this difference between the techniques are that NMR studies the spin of nuclei while ESR does so for unpaired electron, making this a great technique for the study of organic radicals and metals. The only point to make is that the wavelength of light utilised to excite the unpaired electrons is around 30,000,000 nm (3 cm), this technique will not be further discussed as it is outside of the scope of this review. The final technique to discuss is chemi-luminescence[46], which is a
technique quite closely connected to photolysis. As it produces luminescence (light) after a certain chemical reaction. Two molecules are needed for this reaction to occur, firstly luminol (see figure 4), which is an organic molecule that emits light when oxidised and secondly potassium permanganate, which is the oxidising agent. After the chemical reaction, the modified luminol is put into the excited state, after which it will emit a photon to go to the ground state. The detection component comes from looking towards the difference between the baseline (constant emission of light) and when the environment changes, for example with the introduction of PD products. This is a quite compatible technique due to formation of radicals during PD. This can have an impact on the signal perceived by the detector.
2.6 Hyphenation
Due to the complexity of the samples measured (see Chapter 2.1), single dimensional approaches (LC-UV/Vis) are usually insufficient. Due to this complexity, most approaches contain multiple dimensions of separation and/or detection. Although not all techniques are hyphenated, such as the fact that most degradation studies were done with separating degradation and analysis procedures. The terminology used for this is Off-line methods. In such a particular set-up, it means that sections are separated. This could be either the separation of analysis and experimentation (PD) or the separation of the analysis method (e.g. Off-line SPE then LC-UV). In this study when mentioning Off-line approaches, this indicates the separation of analysis and experimentation, in particular, photo-degradation (experimentation). The other side of this are On-line approaches, meaning that everything is connected from analysis, experimentation to detection, although this terminology is also applied to only analysis techniques (e.g. On-line SPE-LC-MS). In this literature review, it refers to the fact that the experimentation (PD) is connected to the analysis technique. Hyphenation can be a complicated matter and can bring many issues with it, these points will be discussed in Chapter 3.
3 Review on on-line analysis & photolysis
Within this main chapter the examined literature will be discussed. As previously mentioned, it will consist of three subsections to get a better overview of the overall information.
3.1 Without separation – on-/off-line
This sub-chapter will consist of studies that do not use any form of a separation technique (e.g. LC) within their main analysis or experimentation approach. This chapter will also include studies that, for example, use Size Exclusion Chromatography (SEC) for the determination of the molecular weight of a sample, as the purpose of the separation is for a molecular property and not for the physical separation of the compounds, it will be included within this chapter. The use of spectrometric techniques (e.g. MS), as the main analysis approach, will also be included within this sub-chapter. The same applies to other various detection techniques, e.g. fluorescence microscopy. This chapter will discuss various studies that apply PD and their analysis of the process, within the constraints mentioned above. The chapter will go into the analysis approach and results of the various studies, and not the actual detailed PD processes, since that is not within the scope of this review.
Testing the effect of light on various kinds of samples is a straightforward matter, if considering an Off-line approach, without separation, for now, the most barebones solution to such an approach consists of a light source and a sample holder. These are usually contained within a closed space, to regulate temperature, humidity, gas composition, ambient light and/or contain the formation of gas due to photo-degradation. When an experiment is done, the sample is moved to the analysis apparatus and analysed. V. G. Borio et al[13] created such a machine and tested this apparatus by analysing the output of the light
source and photo-degradation of three different compounds. Those being, Atenolol, Collagen and an insect repellent. The goal of the study was to create a useable and stable ultraviolet irradiation chamber (UV-chamber) to test the effect of light on the previously mentioned compounds. The analysis was done with a Dispersive NIR Raman spectrometer (no discussed). The use of separation techniques was considered, but due to Raman spectrometry being inexpensive, less time consuming, non-invasive and having reasonable precision, it was used instead. The paper seems to indicate the use of an On-line approach, this however is not specified in the paper. Although the paper does speak of monitoring, there are no results shown of such an approach of analysis, since only results of single (discontinuous, 20 minute) time intervals were used. The UV-chamber could output 2 ranges of radiation, UVA (365 nm) and UVA+UVB (290 nm to 390 nm) radiation, with a detected output of 26.6 W m-2 and 43 W m-2,
respectively, at the sample position. The output of the light source was mapped for both ranges of radiation to look for any inconsistencies that may arise from reflections or other factors that are not accounted for that may increase or decrease the intensity of the signal. It is also of great importance if any comparison studies (with other UV-chambers) are to be performed. By knowing the specific intensity of the signal that was applied to the sample, a more educated comparison can be made with other experiments, which were done on other machines. The study also looked at the output of the laser of the Raman spectrometer, which was 0.25 W, there was no further specification on the intensity of the laser. They found in literature that a minimum output 1 W was necessary for any polymorphic processes to occur, that may modify the Raman signal and lead to an erroneous interpretation of the measurement. No further explanation was given on the nature of the polymorphic processes.
Three compounds from various areas of science were chosen to be tested in the UV-chamber; Atenolol, a -blocker, is a pharmaceutical drug. Collagen, a structural protein, is very important in the investigation of skin aging and other regions of the human body (e.g. bone, tendons). Lastly, N, N-diethyl-m-toluamide, an insect repellent, which gets sprayed out in the “open” air, inside households, or other closed-off environments. These samples from the various environments give a good estimate
for the samples that might be used by such an apparatus. The samples were analysed with a Raman spectrometer, the paper did not specify if this was done inside the UV-chamber. Spectra were taken before and after irradiation and subtracted from each other to look for any specific differences that could lead towards certain structural or molecular alterations. Although V. G. Borio et al reported differences for all 3 compounds, no significant differences could be seen in the results for Atenolol and Collagen (see figures 5 a/b). Only the insect repellent showed any meaningful deviation. Multiple vibrations showed a decrease while one vibration showed an increase in signal (see figure 5 c) after irradiation.
Figure 5: Raman spectra of (A) Atenolol, (B) Collagen and (C) insect repellent. The top (a) and middle (b) lines are the before and after irradiation spectra, respectively. The black line (c) is the difference between spectra a and b.
The study concluded the change in spectrum of the insect repellent to, likely, be degradation of the compound to reactive species that may cause skin damage, but this was merely speculation and no actual follow-up analysis was performed to verify this. The samples in this study are compounds that were used to understand unintentional photo-degradation, where the absorption of light only due to the emission of light by the sun, limiting the range of wavelengths to the range of solar radiation.
Fluorescent labels, on the other hand, have the sole purpose of absorbing and emitting photons. This changes the way the test has to be carried out. One of its most important properties would then be the number of photons it will be able to absorb and how many of those photons would then be emitted (at
an equal or longer wavelength) from the molecule before degradation.
J. C. White et al[39] studied the photo-stability of the most used fluorescent labels and investigated the
number of photons a specific label would be able to absorb and emit before degrading, together with the calculations (see chapter 2), this was called the photo-destruction quantum yield. This is of great importance for, e.g. immunoassays, cell sorting or anti-body investigations. The reason for this particular study was to find ways to get closer to achieve single molecule detection with fluorescence. The approach chosen by J. C. White et al was to let a solution of one specific fluorescent label flow through a tube and flow cell, where light from a lamp would pass through. A detector was positioned at an angle where the fluorescence would be recorded.
Figure 6: Experimental set up
The illustration of the experimental setup positioned the detector at a 45o angle to only detect the
fluorescence and not get any interference from the light source (unabsorbed photons). Although it is drawn at a 45o angle, it is not specified in the paper, the authors most likely assumed this to be common
knowledge for fluorescence. The variables during experimentation were the fluorescent label (+ additives), flow rate of the solution through the cell and the power of the laser. The plots shown below (see figure 7) consist of laser power vs. flow rate (Right) and flow rate vs. relative fluorescence signal for one specific label at different laser powers (Left). All the fluorescence signals were relative, meaning that it was divided by the highest signal, which is at the highest flow rate (no actual flow rates given). The reason for the use of a relative signal is not explained in the paper.
Figure 7:( Left) Relative fluorescence signal versus flow rate of a solution of R-phycoerythrin at various laser powers. (Right) Flow rate for half maximum fluorescence versus laser power, the solid line represents the linear relation.
The right graph was made to verify the linearity of the relation, which was proposed in the calculations. The calculations showed a formula to estimate the number of photons a particular fluorescent molecule would emit before degradation. This is shown in formula 4, whereby Q and ɸ equate to the quantum yield of fluorescence and quantum yield of destruction, respectively, and nf to the total number of
photons a fluorescent molecule can emit.
𝒏𝒇=
𝑸
ɸ (4)
In addition to the laser power and flow rate, for one specific fluorescent label, R-phycoerythrin, additives were added to explore the possibilities for any improvement to occur due to the change in environment. To narrow down the PD processes that occur during illumination, additives were added that eliminate specific molecules from the environment that could cause certain reactions to occur, specifically, oxygen and peroxide radicals, and molecular oxygen, which are known to form/react when irradiated. The plots showed no change in the graph whatsoever, meaning these processes do not occur at a significant rate, or at all. The comparison between the different labels showed that R-phycoerythrin had the highest survival time, and the addition of significant concentrations of n-propyl gallate reduced the degradation rate. Although this study was performed 30 years ago, it still adds crucial information about the survival times of fluorescent labels and the tools to increase this particular property of a label. Although n-propyl gallate did decrease the degradation rate, it also inhibited the signal. This could not be seen in the plots, as only the relative signals were shown. They proposed that the additive quenched the excited state of the label. In this particular setting quenching of the label in counterintuitive, as the label has the purpose of absorbing and emitting photons. J. C. White et al did compare the quenching with the degradation (without additive) and calculated that a factor of 1.7 increase was seen with the addition of the additive, however this comparison is only applicable when the environment can degrade the labels, if not, then the additive will only quench the label and not increase its life time.
A. Torikai et al[38] proposed quenching and photo-degradation processes with their molecule of interest,
Collagen, in which an additive quenched the degradation rate. A. Torikai et al experimented with films made up of Collagen and Collagen with an additive (Vitamin E) to explore certain photo-degradation processes that may occur, together with the photo-stability of the films with and without vitamin E. The films were exposed to poly- and monochromatic light and were analysed by FT-IR, UV and Electron Spin Resonance (ESR) spectrometry, before and after irradiation. Collagen was chosen as the compound of interest due to its abundance in the human body (in tendons, bones, nails, skin and cell walls). Collagen relates to skin aging over time and damage due to high energy solar radiation (UVB range). Although the various analysis techniques give a good picture on the processes that are occurring, they are not enough to completely verify the proposed reaction schemes. Further analysis is required for this, but some educated points were made, that could give a bit of consistency to the proposed reaction schemes. Both the UV and FT-IR difference (before/after irradiation) spectra at various time stamps of irradiation showed an increasing peak height, with increasing irradiation time. The paper does mention ∆O.D. (optical density) at a specific wavelength, although a very confusing terminology, it seems to be the difference in absorbance between before and after irradiation. The wavelength specified is 280 nm, this paper concluded that this increase in signal was due to the formation of Tyrosine from Phenylalanine. This is postulated as to be the first step in the photo aging of animal skin. A slight remark on the conclusion, the increase in signal is with the first-time interval to be around 280 nm, this is also the reason why the paper concluded the formation of Tyrosine, since the absorbance maximum of Tyrosine is at 278 nm, but with longer irradiation times, the peak maximum at 280 nm seems to be red-shifted. This red-shift is also concluded to be because of the formation of Tyrosine. They give no further
explanation on this matter. A shift is also seen in the FT-IR spectra, this was proposed to be the partial breaking of the hydrogen bonds that make up the helix structure in Collagen. The signal around 3300 cm-1 indicates –OH vibrations; it is decreased with longer irradiation times, together with the red-shift.
This was concluded to be the release of moisture. ESR spectra were also taken, the films with and without vitamin E were irradiated and measured. The spectra showed that the films without vitamin E had a much higher peak height. This indicates more radical formation which leads to photo damage, meaning the additive, vitamin E, suppresses radical formation and quenches photo-degradation. Although there was still an increase in signal with longer irradiation times, this was a much lower signal than without the additive. They concluded that vitamin E does supress the photo aging of Collagen significantly when irradiated. They proposed 2 quenching pathways. One, suggesting vitamin E is an UV-absorber, which means vitamin E competes with Collagen for the light that is passed through the film, reducing the absorption of light. Another proposition would be, that vitamin E is a radical scavenger, meaning, that the formed radicals are transferred to vitamin E instead of the Collagen degrading. Both are viable suggestions. No further experimentation was shown to conclude whether it was one or the other, or if it is another process all together. The paper also investigated the possibility of thermal-induced changes. The FTIR and UV spectra showed (after heating) a reduction in the OH- band, and increase in peptide linkage, respectively. This result showed that Collagen, although having the property to retain water, loses it, when heated. This gave more insight into the structure of Collagen and how heat affects it.
The results discussed up till now only looked at the changes that occurred directly on the molecule of interest. Another way of investigating PD on specific compounds is to measure the changes indirectly, as Fechine et al[21] did. They investigated the photo-stabilisation of poly (ethylene terephthalate) (PET),
by investigating carbon dioxide (CO2) formation using FTIR spectrometry. The intensity of the FTIR
signal before, during and after UV-exposure was measured to determine the CO2 that formed, relatively,
since only absorbance units were compared. Many aspects were investigated: mono/multi layered polymers, effect of accelerated UV-exposure, different environments and the impact of UV-absorbers in/on the polymer. Although this was done in an On-line manner, measurements were done in 20 minute intervals. Spectra were not shown, only absorbance units at unspecified wavenumbers. It is not mentioned if the cumulative signal was taken from the complete IR-spectrum or if only specific wavenumbers were chosen (e.g. ~2350 cm-1). This is of importance, as this could impact the results
achieved, as the signal shown could have come from multiple sources (instead of only CO2). Although
this could be a big issue, it is not since the PD of PET has already been heavily investigated by the authors. The goal of the study was to investigate the photo-stabilisation considering different environments, this is of great importance for the prediction of life times of PET. This has consequences for, e.g. plastic wrapping around foods or plastic containers. In water bottles for example, UV-stabilizers, if used improperly, could dissolve into the water and be consumed by the user, which could cause frightful health hazards. From their investigation in literature of PD on PET, before experimentation, they found that, although PD was observed on the surface, less degradation was observed on the backside, and little to no degradation was observed in the interior. The reason for the lack of degradation in the interior was due to air starvation. This showed that PD was dependent on air, not completely, as will be seen from experimentation discussed later. Although PET heavily absorbs below 315 nm wavelengths of light, the longer wavelengths of light that are not absorbed will pass through the surface and reach the backside (see figure 8).
Figure 8: Illustration of the side view of a polymer film, portraying the positioning of oxygen and its relation to PD. From this it could be determined that wavelengths of light above 315 nm are capable of inferring degradation, due to these conclusions, one could determine that the area in need of (UV-) “protection” is the surface, and protecting it from most wavelengths of light (above and below 315 nm) and inhibiting degradation at the backside, together with removal of air, little to no degradation should be observed. Four different samples (films) were investigated. The first two consisted of a mono-layer of polymer, one with and the other without UV-absorber (1% of total). The last two samples consisted of multi-layered films, which were multi-layered on top of each other in the form of A/B/A/C, where: A: PET with x% of UV absorber, 0.7 µm thickness; B: PET without UV absorber, 10 µm thickness; C: PET co-polyester without absorber, 0.6 µm thickness; x is either 0.23% or 0.93% UV-absorber. The reasoning behind this particular setup was not specified, only the fact that degradation was observed on the surface (front and backside), thus applying the UV-absorber on those sides. The reasoning behind the PET co-polyester film, is most likely due to the fact that there would be no direct interaction between the UV-absorber and the contents inside a particular wrapping or container (e.g. bottle). They do not explain this in the paper, this was deduced from the issues that came with PD of PET and the use of UV-absorbers. The films were placed in a reaction cell, where UV-light shone perpendicular on the films, while the IR beam went parallel and beside the film, to capture the spectrum of the gas. A Xenon lamp was used, together with an AM1.5 filter, to cut-off the wavelengths of anything below 300 nm. For every experiment, a two-hour exposure time was taken. The cell was sealed, letting neither any gas in or out. There is a reference to another study mentioned, where the reaction cell is described in more detail. The environment of the cell was varied, using three different kinds of gas: oxygen, with (wet) and without (dry) water, and (dry) nitrogen gas. There is no mention of temperature control, which is actually very vital to PD experiments, as this impacts the (PD) reaction greatly. There is also no mention of light intensity of either the IR- or UV-beam. Although these are not mentioned about the reaction cell used for the main experiments, the light intensity and temperature is mentioned in the UV pre-conditioning experiments. These experiments consisted of exposing the samples to 9 days of UV-light before the main experiment, to determine if “auto-acceleration” occurred, as it is called in the paper. Auto-acceleration, meaning that the polymer is exposed to an aggressive environment for a considerable amount of time and the reaction products formed during this prolonged exposure take part in the main experiment to accelerate the overall degradation. This study mentions that this is common practice to investigate this process in the laboratory. The exposure of the pre-conditioning experiments was done on different equipment. The UV range was matched with that of terrestrial solar radiation in the wavelength range up to ~360 nm, as mentioned in the study. The intensity of the light below 320 nm was 2.0 ±0.3 W/m2, there is no mention of the range above 320 nm. The temperature was at 30 ±1 oC
Continuing on with the measurements done, the first one, was measuring the CO2 formed from the 4
samples without pre-exposure and only wet oxygen was used (see figure 9).
Figure 9: Absorption at 2350 cm-1 analysed at specific times for various PET films, showing the increased signal over time,
indicating PD.
As can be seen in the graph, when the UV-light is turned on an increase in signal is observed. The rate at which the signal increases differs slightly between PET and PET1/2/8. The difference is the addition of the UV-absorber to PET1/2/8. This graph shows that with the addition of an UV-absorber the PD decreases, in this case by ~25 %. From this one can also conclude that using only 0.23 % UV-absorber on the surface, gives the same effect as using 1 % UV-absorber mixed with the polymer. Using 0.93 % UV-absorber on the surface gives the same effect as using 0.23 %, meaning that there is a saturation limit. Unfortunately, they did not do an experiment mixing only 0.23 % UV-absorber with the polymer (creating a mono-layer) to see if the saturation limit was not already achieved at that percentage. They continued with the measurements under different environments and observed that dry oxygen and nitrogen gave the same results, pointing towards only direct PD reactions, while wet oxygen gave an increased signal, from this they concluded that water accelerated the PD. This did not follow the results from earlier work done by Wiles and co-workers, which is mentioned (not shown) in the paper, as Wiles and co-workers suggested a reaction scheme that involved oxygen instead of water. The study concluded that this was due to the short exposure times (2 hours). The pre-exposure samples were also investigated, they exposed 5 PET samples (3 with wet oxygen, 1 with dry nitrogen and oxygen) to the UV-light after pre-exposure of 9 days. The results showed that the wet oxygen samples were not very repeatable, this was concluded to be due to the minor inconsistencies in the structure of the different films that were magnified due to pre-exposure. An increased signal was observed for the dry oxygen environment compared to nitrogen, which coincided with the schemes proposed by Wiles and co-workers. The dry nitrogen environment showed a 50 % decrease in signal compared to the samples without pre-exposure. It was concluded to be due to the starvation of oxygen during the pre-exposure period. The dry nitrogen and wet oxygen environments on PET1/2/8 were also compared. The results observed were expected, as both decreased around the same value (around 20 *10-4 absorbance units)
for dry nitrogen as well as for wet oxygen. In conclusion, the UV-absorber does increase the photo-stability of PET, and that applying 1/5th of the substance on the surface instead of mixing it with PET
provides the same “protection”. The environment also greatly impacts the PD reactions, as it supplies a stream of molecules that give it room to react through different mechanisms (indirectly). Jin et al[23] did
of TiO2 (Anatase/Rutile) with various treatments (Sulphate/Chloride) and coatings (Al2O3/SiO2), as
UV-absorbers on polyethylene (PE). Even though the term UV-absorber is used, most of the coated samples did not in fact reduce the PD of PE. TiO2 is known for its catalytic PD, as it forms hydroxide
radicals due to UV-exposure, depending on the many variables that come into play (e.g. UV-filters, atmosphere). Similar to G. J. M. Fechine et al, they experimented under different conditions, varying the environment (gas around sample), however, this particular study did vary the intensity of the UV-exposure as well. The reaction cell was very similar as the previous study. The IR-beam passes through the reaction cell parallel to the PE film and the UV-light is passed through the film, perpendicular to the IR-beam. The set-up was not shown in the previous study, the reaction cell in this study is illustrated in figure 10.
Figure 10: Front and side view of the sample holder and IR/UV Chamber.
The study of C. Jin et al described the experimental parameters in slightly more detail, as they mentioned the power and range of their UV-source (150 Watt, 280 – 800 nm, respectively), for PD, however, this is the power and output before passing through the filters (AM0/AM1.5). They did show a UV-spectrum of the output, but this only specified the wavelength range and used arbitrary units on the y-axes, the UV-spectrum is shown in figure 11.On the x-axes, µm is mentioned, this is an error, as in the paper (the text) it is described as nm.
Figure 11: Irradiation experienced by the sample through various filters.
During the experimentation both the AM0 and AM1.5 filter was applied, due to its cut-off <300 nm. A 100-mm water filter was set up as well, to reduce IR heating. The results observed will not be discussed in detail, as most of the measurements came to the same conclusion as the previous study. Only the notable results will be mentioned and due to the UV-lamp drifting time (not further specified by the
author) over, the results which had a long-time delay between experiments were not compared. One of the peculiar results observed was the accelerated rate of the signal with the long exposure experiment (1400 hours).
Figure 12: (a) FTIR transmission spectra of carbonyl groups for samples pigmented with Anatase (A2) or Rutile (R3,R4,R5) exposed to UVA-340 for ~1100 hours. (b) Carbonyl absorbance of pigmented polymer films with irradiation from UVA-340. As seen in figure 12, the signal (b) seems to accelerate, however this could also be within the error, since no duplicates were measure, nothing can be said about this result. The study only mentions the fact that the order of the sample coincides with the theory and experiments done during short exposure times (3 hours). It is most likely not within the scope of the project, and it was not of interest to investigate further. Figure 12 a shows that the intensity of the IR-spectrum (at 1710 cm-1) is increased,
pointing heavily towards CO2 formation. The UV-intensity versus the CO2 absorbance was also
investigated, for unpigmented and Anatase pigmented PE samples (pigmented meaning with TiO2). Of
which the results are shown in figure 13.
Figure 13: Unpigmented (a) and pigmented (b) polymer films after 240 min of irradiation. Showing the linear dependence of CO2 on the UV intensity / square root of the Transmission.
Although the results show expected linearity (a), the UV intensity is shown in arbitrary units. The actual numbers of the UV intensity are not explained and if it is comparable between the different experiments (a/b), however the regulation of the intensity is changed with the use of neutral density filters, for both
(a/b) experiments. The pigmented sample was radiated with the AM0 and AM1.5 filters, while the unpigmented sample was not. The reason for the lack of filters with the unpigmented sample was done for maximum sensitivity, while the filters were used with the pigmented sample without explanation. The square root is taken of the transmission, as this gives linearity. The square root relation is due to the fact that the most predominant CO2 formation is due to the catalytic PD with TiO2, as the metal
absorb highly at the wavelengths emitted from the UV-source, due to this absorption the PD reactions will be dominated by the metal, causing the relation between it and CO2 formation to be linear with the
square root. This relation comes from formula 5, whereby K and I are the absorption coefficient and intensity of UV radiation, and ksO and kr are the rate constants of the surface formation of hydroxyl
radicals and recombination of electron and holes, respectively. The formula is further elaborated in [23].
𝑹𝒂𝒕𝒆 𝒐𝒇 𝒇𝒐𝒓𝒎𝒂𝒕𝒊𝒐𝒏 𝒐𝒇 𝒉𝒚𝒅𝒓𝒐𝒙𝒚𝒍 𝒓𝒂𝒅𝒊𝒄𝒂𝒍𝒔 = 𝒌𝒔𝑶∗ √
𝑲𝑰 𝒌𝑹
(5)
The TiO2 acts as a catalyst for the formation of OH radicals, which in turn react with PE to form CO2.
With the formation of hydroxyl radicals, oxygen and water are necessary, this can be seen in figure 14.
Figure 14: CO2 analysis of A2 pigmented polymer films under various gas compositions.
Applying an environment with only nitrogen gas and water (wet) gives only a minor increase in CO2
formation, while increasing the oxygen content with only 20%, increases the signal immensely. The PD with the wet nitrogen environment was not fully investigated. They hypothesised it to be the degradation of TiO2, as a greyish colour was observed after the experiment. It was not further examined. The goal
of this study was to develop a method to examine PD of polymers within reasonable times (hours instead of weeks). This was achieved; however some minor hindrances were seen (e.g. not comparable results, due to time drift). This leaves the conclusions with some ambiguity. The study does inform the reader that these are preliminary results and more investigation is needed. The TiO2 used by C. Jin et al was
Anatase and Rutile, Anatase specifically caused considerably more PD to happen to the film, aiding it, instead of inhibiting it. This was the reason why more samples consisted of Rutile, as the goal of this study was to increase photo-stability.
Alberici et al[37] had the opposite goal, as they investigated the photo-catalytic degradation of
Chlorinated Volatile Organic Compounds (CVOCs) with TiO2 and UV-light. The analytical tool used
for this particular study was Mass Spectrometry. Alberici et al used an On-line monitoring approach to follow all the processes. MS2 was also applied for the structural characterisation of the products. The
together with the abundance of the compounds in various areas in society, mainly in gaseous emissions from industrial processes, landfills, hazardous waste sites, wastewater purification facilities and indoor air (not shown). The need for fast, effective and inexpensive techniques of the destruction of CVOCs is quite demanding. This study investigated the potential of TiO2 combined with UV-radiation, looking
for signs of complete degradation and mineralisation (the complete oxidation of a particular compound (e.g. CHCl3 CO2 + HCl + Cl2)). The CVOCs of interest consisted of trichloroethylene (TCE),
tetrachloroethylene (TeCE), chloroform and dichloromethane. Their respective by-products and final products were phosgene, dichloroacetyl chloride (DCAC), trichloroacetyl chloride (TCAC), molecular chloride (Cl2) and carbon dioxide (CO2). The set-up of the on-line monitoring approach is illustrated in
figure 15.
Figure 15: Schematic representation of the MS/MS2 on-line monitoring photocatalytic degradation system. A) UV/VIS-Lamp; B) Photo-reactor; C) VOC reservoir; D) Flow-meter; E) Carrier gas; F) Exit line; G) Needle valve flow control, H) MS; I) Ion source; J) Detector; K) Light starter; Qn) Quadrupole analysers, qn) Collision cells. Items are not shown to scale. Starting with the valves at the most right (G, right-side), the carrier gas (N2:O2 – 80:20 ratio, with either
20 % or 80 % relative humidity) flowed in from E through the flow control valves (G, right-side). One tube goes directly to the mixing T (D), while the other goes through Cinlet, which contains the CVOCs.
A standard of 500 ppm (parts per million by volume) was prepared beforehand. To keep the concentration consistent even at varying (intentional) flowrates, the flow of gas was altered by adjusting the flow into the mixing T. At B, the photo reactor is positioned, inside it, the carrier gas (with the CVOCs) flows passed the TiO2 coated wall. The box at K is described at the light started, one can
assume this to be the UV-source. The light (365 nm, 30 W) passes through the photo-reactor, akin to the carrier gas. The outlet at F is described as the exit line, presumably used as a redirection of the gas, when the MS is not in use. The valve flow control, G (left side), connects the PD experiments with the analysis technique (MS). H is described as the MS, I and J being the ion-source and detector, respectively. Qn are the quadrupole analysers, while qn are the collision cells. The MS used election
ionisation (EI) and applied multiple selective ion monitoring (SIM). The major ions in the mass spectra (70 eV) are shown in the table below. The ions selected for SIM are the most abundant in the mass spectra (see table 1).
Table 1:Major ion observed in the 70 eV EI mass spectra of selected compounds.
The selected m/z values are for: TCE – 95, phosgene – 63, DCAC – 83, CO2 – 44, TeCE – 166, TCAC
– 117, chloroform – 83, dichloromethane – 49. MS2 collision-induced dissociation (15eV, Argon, CID)
experiments were utilised, for the structural characterisation of the by-products and final products, when necessary.
The study initiated with the degradation of TCE. The mass spectra at specific intervals showed a decrease in TCE signal, and was no longer visible at t = 30 minutes (see figure 16).
Figure 16: MS Spectra of TCE at specfic times (a) 0min; b) 2min; c) 30min) during the PD process
The characteristic chloride isotope ratios are visible (35/37Cl, 3:1 ratio). The spectra suggest that TCE
was completely mineralised to CO2, with Cl2, phosgene and DCAC as by-products, meaning that with
present conditions, 20 % humidity, 1000 mL min-1 flow rate and a TiO
2 coated wall, give rise to rapid
PD. MS2 CID was applied as well, the expected molecular structures were observed (see figure 17).
Figure 17: Collision induced dissociation product spectra, whereby the molecular weight in the black balls is isolated in the quadrupole and is hit with argon to generate collision, and in turn dissociation. Four (a-d) different ions were selected and
CID was applied.
35Cl and 37Cl of each compound were selected, to obtain more information on the molecular structure
of the molecule. The observations were expected, with no results out of the ordinary. The on-line monitoring results were no different (see figure 18).
Figure 18: On-line monitoring of TCE after PD process is initiated, at three different speeds (a-c; 100, 500 and 1000 mL min -1).
Three different flow rates were applied, 100(a), 500(b) and 1000(c) mL min-1, due to these different
rates, the residence time inside the reactor will also be shortened (at higher flow rates). Graphs A and B show no difference (excluding intensity), while graph c shows the formation of a third component. This was concluded to be an intermediate to the production of phosgene, giving the intermediate component no time to continue the reaction. After some time, the three graphs reach a steady state, meaning that after some stabilisation, no further changes are observed in the experiment. The humidity of the specific experiment was 20 %, the mass spectrum (figure 19) shows the steady state with the humidity at 80 % and flow rate at 1000 ml min-1.
Figure 19: MS spectrum of TCE after 30 minutes of UV irradiation, at a speed of 1000mL min-1 and 80 % humidity.
The expected m/z values (see table 1): 95 (for TCE), 63 (for phosgene), 44 (for CO2) are observed. An
increase, compared to 20 % humidity, in TCE is observed, together with the lack of DCAC formation. The authors speculated that the retardation of the TiO2/UV photocatalytic degradation was due to the
The study referred to Nimlos et al [not shown], who proposed the illustrated scheme:
Scheme 2: Proposed pathway for TiO2 assisted PD of TCE
The OH-radicals, formed due to the reaction of water with TiO2/UV, react with TCE, causing it to lose
a chloride atom, forming a radical molecule, which can react with O2 to form the molecule located at
the centre. This molecule can either lose a chloride radical or a Cl2CH-radical to form the by product.
Phosgene can further split to form CO2 and Cl2. The ratio between reactions (a) and (b) depend on the
relative bond strengths of C-Cl and C-C. The highlighted molecules proposed by Nimlos et al have been observed during experimentation.
The TeCE, Chloroform and dichloromethane will not be discussed, as the results are very similar, adding no new information. The photo degradation rates were compared, qualitatively, as no actual numbers are shown. The results were compared with other studies, as agreed with each other. Resulting in the order: TCE ≥ TeCE > dichloromethane > chloroform. The goal of this study was to observe the applicability of on-line MS and MS2 techniques for reliable monitoring and characterisation of the
products of TiO2/UV photo-catalytic degradation of CVOCs. This was accomplished, together with the
determination of the potential of TiO2, as a fast and reliable treatment of CVOCs.
Shifting to another area of science, where TiO2 is of interest as well, this being conservation science,
the research of TiO2 is of great importance, as many great artists (e.g. Picasso, Pollock and Mondrian)
have used this particular white pigment in their respective paintings. It is known that a particular form of TiO2 (Anatase, powder) is capable of acting as a photo-catalyst while an another form of TiO2 (Rutile)
is less or not active at all, as discussed before, because of this, these particular paintings could degrade because of this pigment, as such, conservation scientists are appropriately interested in this pigment. One of the important factors of conservation is prevention of (further) degradation on any piece of art, be it a painting, sculpture or any other form of art, thus van Driel et al[3] attempted to determine early
warning signs for photo-catalytic degradation of TiO2 by means of surface analysis, on white oil
paintings in particular. The determination of degradation of white oil paintings now is the detection of chalking which can be seen with the naked eye. The degradation has at that point become visible,
changing the actual appearance of said painting. Chalking is the formation of a white powder on the surface, this white powder is the TiO2 that has risen to the surface. An oil painting that has not degraded
will have oil on the surface (in this particular case, linseed oil), while the TiO2 is beneath this oil. This
oil substance will degrade due to the creation of radicals. The radicals are formed because TiO2 absorbs
light, and reacts with its surrounding. The radicals degrade the oil to form volatile compounds, such as CO2 and H2O. The oil on the surface will decrease, letting the TiO2 surface, thus creating a white powder
on the surface. The kind of degradation will fade the original colour of the painting, as such, preventing such event to happen increases the lifetime of such paintings. This is done by moving the painting, if early signs of degradation are seen, to a dark room, halting the degradation process, till a proper solution is found, such as reapplying a new coat of oil on the surface, or removal of radicals or air, as these are the causes of PD. This however will only work if the degradation is found before reaching the chalking stage, preferably before the surfacing of TiO2, as the discolouration will be minimal (keeping the
original appearance of the painting.
B.A. van Driel et al proposed the use of surface analysis techniques. The techniques proposed are AFM and XPS together with gloss measurements and inspection with the visible eye to observe chalking. The gloss measurement will not be explained in detail as this is not within the scope of the project, see chapter 2.5 for AFM and XPS. The gloss measurement is performed at specific angles (85o), and gives
a corresponding gloss number at specific points, a higher value meaning glossier. The amount of light that is reflected from the surface at specific angles determines the gloss number. A higher number is given if the reflection is mirror like, meaning less random (smooth surface). If the surface is rough the reflections will go in random directions, this is called matte (opposite of gloss). The three surface analysis approaches together determine the progress of the PD process, starting at an initial not degraded stage (1) to an oil breakdown stage (2), advancing to TiO2 surfacing stage (3) and
ending with a chalked paint stage (4), which is visible to the eye. These stages are illustrated in figure 20. The figure depicts a simplified version of a white oil painting, where the TiO2 particles are positioned
below the oil layer, as discussed before. The oil layer on top is called the medium skin. B. A. van Driel et al proposed the following processes, depicted below.
Figure 20: Proposed model of linseed oil degradataion mechnism in relation to gloss, AFM and XPS analysis.
The TiO2 particles generate radicals which interact with the oil layer that would decrease the medium
skin (2). With further degradation, the TiO2 particle will be protruding from the surface (3). This will
